Impact of chemotherapy-induced amenorrhea in breast cancer patients: the evaluation of ovarian function by menstrual history and hormonal levels

Breast cancer accounts for 28% of newly diagnosed malignancies among women [1], and almost 25% of those in Western countries occur in premenopausal females [2]. Conversely, the majority of newly diagnosed breast cancers in China are premenopausal females (with age ranging from 40 to 49 years) [3]. As more breast cancer patients benefit from the widespread use of adjuvant chemotherapy [4], long-term side effects, particularly ovarian suppression, become a major concern among women of reproductive age [5]. Ovarian function assessment after chemotherapy in premenopausal or perimenopausal women with hormone-receptor-positive breast cancers is critical in clinical decision making, particularly in cases involving primary adjuvant endocrine therapy. More than one-half (52.2 to 77.7%) of the number of premenopausal women experience chemotherapy-induced amenorrhea (CIA) [6, 7], which results in infertility, psychological distress, and prolonged exposure to risks of menopause.

Patient age, intensity of chemotherapy, and tamoxifen dosage are primary determinants of CIA. Older women (>40 years) have a higher risk of developing amenorrhea after adjuvant chemotherapy [8‐10]. Previous reports indicated the relevance between currently used chemotherapy regimens (anthracycline-based, taxane-based, and cyclophosphamide/methotrexate/fluorouracil (CMF) regimens) and the onset of CIA [9, 11‐14]. The occurrence of amenorrhea is frequently associated with the cumulative dosage of chemotherapy [12, 14]. The effect of tamoxifen on menstrual cessation has also been determined in several multi-central randomized controlled clinical trials [13, 15, 16]. However, most of the evidence comes from women in Western countries. Evidence of the occurrence and characteristic factors of CIA among Asian women remains insufficient.

In this study, we conducted a retrospective analysis to assess the onset of CIA and menstruation resumption. Potential factors relevant to CIA and menstruation resumption were determined by employing univariate analysis. Differences in the evaluation efficacy of ovarian function between two methods (menstrual history and hormonal levels) were also discussed.

CIA is one of the frequent therapy-related adverse events observed in patients who undergo chemotherapy. CIA is a significant drain on the quality of life of cancer patients because of its associated menopausal symptoms and potential risk of infertility. In this study, we conducted a retrospective analysis to assess the onset of CIA and menstruation transformation among Chinese women who underwent chemotherapy for breast cancer. A total of 73 premenopausal breast cancer patients were included in this study, whereas 295 patients were excluded for various reasons. The major clinical characteristics in the included group (73 patients) and in the excluded group (295 patients) were almost the same. Although only 19.8% of the potential group is included in our analysis, the potential selection bias is limited.

In this study, the rate of CIA (83.6%) was similar to the data provided by other trials. In the International Breast Cancer Study Group trial 13–93 [17], the general incidence of CIA was 86%, which is approximately equal to our results. Han and colleagues reported that the rate of CIA in the Korean population was 89.1% [13]. However, the incidence of CIA was lower (68%) in the findings of Lee and colleagues [16]. The differences may have resulted from the different age compositions among these studies. The higher component ratio of included patients who were ≥40 years old (78.1% in our study versus 61% in the study of Lee and colleagues) may be attributed to the higher rate of CIA onset in our study.

Previous evidence indicates the incidence of CIA depends on the patient age or type and intensity of chemotherapy or tamoxifen dosage [18]. In this study, we also discussed these potential factors on the onset of CIA and the transformation to menopause via univariate analysis. Our study demonstrated that age at diagnosis is an important influencing factor on the onset of CIA, which is in accordance with previous studies [7, 17, 19‐22]. Age at diagnosis was also a major predictor for both reversible amenorrhea and menopause, as shown in Additional file 1 Table S4. Additionally, we found that Chinese women younger than 40 years who experienced CIA had higher incidence of menstruation resumption. Similar results were reported in other Asian-based studies [13, 16], as well as in Western studies [6, 22]. Patients aged ≤45 years old were less likely to develop menopause after CIA and more likely to experience menstruation resumption or maintain amenorrhea, as confirmed by both clinical manifestation and serum hormone status in the study.

Previous evidence indicates that tamoxifen is a potential influencing factor of CIA onset. In this study, the relevance between the administration of tamoxifen and CIA was not significant in the univariate analysis. However, this relevance was influenced by the age of the patient at diagnosis [20, 22]. Given the limited number of patients in different age-based subgroups, conducting subgroup analysis, as well as determining the association between the tamoxifen and patient age, is difficult. Our data indicated that the addition of tamoxifen substantially affected the onset of menopause. The possible explanations are the disturbance of negative feedback from the hypothalamus to the ovary [18] and the depression of pituitary production triggered by tamoxifen [23, 24].

Menstrual status determined by the history of menstrual bleeding is a conventional and convenient method of assessment in determining ovarian function [9]. However, menstrual status cannot thoroughly reflect actual ovarian function [25]. The serial measurement of serum ovarian biochemical markers is an acceptable method to reflect ovarian function. Accordingly, in this study, we determined patient menstrual status by employing both clinical menstrual history and serum ovarian biochemical markers. We compared the differences in the median time to onset of permanent amenorrhea between the aforementioned two approaches. Our results indicate that the time of clinical cessation of menses was earlier than the recorded serum postmenopausal hormonal levels. As shown in our study, for Asian patients aged between 46 and 49 years, another 21 months were needed for hormone levels to reach the postmenopausal range after the cessation of menses. For patients older than 50 years, another 17 months were necessary. The World Health Organization defined 12 months of amenorrhea as an indication of menopause [26]. On the basis of our results, this criterion is not reasonable for identifying women who experienced premature amenorrhea after cytotoxic therapy. Accordingly, the serial measurement of serum FSH and E2 combined with menstrual history are suggested as an assessment for ovarian function [27]. Given the complexity of ovarian change after chemotherapy, the duration of hormone measurement after amenorrhea onset for postmenopausal status confirmation is uncertain. The guidelines provided by the National Comprehensive Cancer Network suggest a postmenopausal range of 12 months to meet the criteria for menopause. However, most studies support the regular monitoring of hormone status without mentioning the duration [27, 28]. We propose that the clinical history and serum hormone level follow-up duration should last for at least 21 months for patients aged 45 years to 49 years and 17 months for patients aged ≥50 years to determine the exact menstrual status among Asian women.

Previous studies preferred FSH as a biochemical marker to predict ovarian reserve [26]. The role of FSH in accessing ovarian function has been well demonstrated. Other than FSH, anti-Müllerian hormone (AMH) and inhibin B are also shown to improve the predictive capacity of ovarian function [29, 30]. AMH and inhibin B appear to be superior to FSH because they are not affected by the administration of tamoxifen [30]. Moreover, AMH and inhibin B reflect subtle changes in menstrual transition, which has been compared with FSH. AMH is also strongly recommended for its sensitivity in predicting ovarian function and stable expression over the menstrual cycle [25]. With regard to biophysical parameters, antral follicle count (AFC) is also a useful parameter. The function of AFC in assessing ovarian function among children cancer survivors has been proven [31]. Reduced AFC was observed in breast cancer patients after chemotherapy and appeared to be in accordance with suppressed ovarian function [25]. Further investigation is necessary to confirm the function of such biomarkers in clinical application.

Our study had several limitations, such as a short follow-up period and small sample size. As previously stated, the large proportion of patients dropping out of the study was assumed to have resulted from the lack of periodic hormone values. Approximately 26% hormone insensitivity and poor adherence would be the possible reasons for the missing information, which introduced bias. The small number for certain groups (for example, patients without amenorrhea) can affect the precision of CIA rates in perimenopausal breast cancers. As a retrospective study, the menstrual status might be less reliable due to a recall bias. Despite these limitations, our study is an effective exploration of time points and intervals between them to predict the ovarian function pattern during follow-up. A more comprehensive study with a longer follow-up period and written diaries of patients is planned by our institution in this area. In addition, other hormone markers, including AMH and inhibin B, could be assessed serially in determining post-chemotherapy ovarian reserve, and the related cutoff point for these biomarkers could be evaluated.

An accurate and probable evaluation of ovarian function promotes better understanding of the effect of cytotoxic agents, thereby offering cancer patients useful information with respect to appropriate endocrine therapy and fertility preservation.